Continuum Research Methods Series Research Questions - Richard Andrews
Evaluation Methods in Research — Judith Bennett
Analysing Media Texts - Andrew Burn and David Parker Action Research - Patrick Costello
Ethics in Research - Ian Gregory
Researching Post-compulsory Education — Jill Jameson and Yvonne Hillier
Systematic Reviews - Carol Torgerson
Using Focus Groups in Research - Lia Litosseliti Real World Research Series
Developing a Questionnaire - Bill Gillham The Research Interview - Bill Gillham Case Study Research Methods — Bill Gillham
Analysis
Terry Locke
continuum
Continuum International Publishing Group The Tower Building 15 East 26th Street 11 York Road New York London SE1 7NX NY 10010 © Terry Locke 2004
All rights reserved. No part of this publication may be reproduced or trans-mitted in any form or by any means, electronic or mechanical, including photocopying, recording, or any information storage or retrieval system, with-out prior permission in writing from the publishers.
British Library Cataloguing-in-Publication Data
A catalogue record for this book is available from the British Library. ISBN 0 8264 6486 6 (paperback)
Typeset by YHT Ltd, London
show which space of freedom we can still enjoy and how many changes can still be made.
Acknowledgements viii 1 A Close-up on Text 1 2 Language, Discourse and Context 11 3 The Critical Turn: Making Discourse Analysis
Critical 25 4 The Question of Metalanguage in CDA 40 5 Analysing a Print Text 54 6 Analysing Oral Texts 74 References 90 Index 94
Acknowledgements
My first debt of gratitude is to those thinkers about
lang-uage and its place in the world whose pioneering work
made this book possible, in particular, Michel Foucault,
Mikhail Bakhtin, Norman Fairclough, James Gee, Ruth
Wodak and Teun van Dijk.
I would like to thank the New Zealand Herald for
per-mission to publish the editorial that appears in Chapter 5.
I am also grateful to the generosity of Oneroa Stewart
who, in a number of emails, made it possible for me to
contextualize this editorial. My thanks go to Lester
Flockton and Professor Terry Crooks, for permission to
use NEMP data for the transcript which is reproduced in
Chapter 6. I would like to thank Michael Smither for
permission to reproduce his painting 'Gifts' (Chapter 6)
and the Collection of Museum of New Zealand: Te Papa
Tangarewa, as owner of this work. My thanks also go to my
colleague Graham Price who helped with the transcript
used in Chapter 6.
1
A Close-up on Text
Critical discourse analysis in a nutshell
One of the founders of critical discourse analysis (CDA), Norman Fairclough, has described it as aiming
to systematically explore often opaque relationships of causality and determination between (a) discursive prac-tices, events and texts, and (b) wider social and cultural structures, relations and processes; to investigate how such practices, events and texts arise out of and are ideologically shaped by relations of power and struggles over power. (1995, p. 132)
Summed up in a number of bullet points, CD A:
• views a prevailing social order as historically situated and therefore relative, socially constructed and changeable.
• views a prevailing social order and social processes as constituted and sustained less by the will of individuals than by the pervasiveness of particular constructions or versions of reality - often referred to as discourses. • views discourse as coloured by and productive of
ideology (however 'ideology' is conceptualized). • views power in society not so much as imposed on
individual subjects as an inevitable effect of a way par-ticular discursive configurations or arrangements
Critical Discourse Analysis
privilege the status and positions of some people over others.
• views human subjectivity as at least in part constructed or inscribed by discourse, and discourse as manifested in the various ways people are and enact the sorts of people they are.
• views reality as textually and intertextually mediated via verbal and non-verbal language systems, and texts as sites for both the inculcation and the contestation of discourses.
• views the systematic analysis and interpretation of texts as potentially revelatory of ways in which discourses consolidate power and colonize human subjects through often covert position calls. (For other precise accounts see Fairclough and Wodak 1997, Janks 1997, Wodak (1996, 2001).)
This book is about critical discourse analysis as a research method. However, CDA might be better described as a scholarly orientation with the potential to transform the modus operandi of a range of research methodologies. In respect of educational research, it has the potential to reveal the way power is diffused through the prevalence of various discourses throughout an education system, at both the micro-level of individual classrooms and the macro-level of large-scale reform. As in other settings, CDA has to be seen as a political intervention with its own socially transformative agenda.
A close-up
Some years ago, a billboard appeared near my home in Kingsland, Auckland. Against a white background were painted with brush-like, black strokes the words: 'Kelly
Browne's parents are away. PARTY at her place!!' (see Figure 1.1).
Figure 1.1 Kelly Browne Billboardard
This is a text of just nine words and two sentences. A linguist might comment on the syntax by remarking that the first sentence is simple with a subject and one finite verb, and the second is either a noun phrase (a group of words built on the noun headword 'party') or a simple imperative sentence (where 'party' functions as an imperative verb telling people to party at Kelly's place). Such comments clearly utilize specialized linguistic vocab-ulary and will make sense to someone familiar with terms such as simple sentence, subject, finite verb, noun phrase and imperative. As a reader, you may find this linguistic infor-mation interesting, or you may be thinking, 'Do I need to know this?'!
There is, however, another way of approaching this text, which focuses more directly on the broad question, 'What does it mean?' To explore this question, I'd invite you to engage with a task I have used with a number of my students. To begin exploring the relationship between
Critical Discourse Analysis
text and what I will be calling discourse, I show these
students a reproduction of this billboard (see Figure 1.1).
I point out that the text is made up of words, but that the
writing has been done in a particular way (rather like
graffiti). In doing so, I am introducing the idea that there
are three types of signs: icon (where the sign resembles its
object in some way, as in road signs); index (based on
association or causality, e.g. graffiti signifying anti-social
behaviour); and symbol (where the connection is
convent-ional, e.g. a rose signifying love) (Hodge and Kress 1988:
21-3). I invite students to invent a 'story' that explains the
connection between the first statement and the second.
As an educator of teachers and a researcher with an
interest in language, what is interesting to me is the
degree of consensus arrived at by the frequently diverse
students involved. The consensus story might be told
thus:
Kelly Browne is in her late teens or early twenties and is still living at home with her parents. She is past the age where she feels the need to check out what she does with her parents. She likes to party, but can't imagine a party as being an occasion that includes parents. Parents, by defin-ition, are simply not party animals and tend to be some-what restrictive. Parties - real parties that is - are more about breaking out (and sometimes taboos and furniture). In general, then, a house without parents is an ideal locat-ion for a party.
Clearly, this consensus story has far more detail that the
nine-word billboard. Despite the presence of dissenting
voices, however (more on this below), it constitutes a
coherent, widely subscribed to, interpretation of the
billboard. Two questions immediately present themselves;
'Whence did this story originate?' and 'What is meant by
the word interpretation? These questions will turn out to be big ones and central to the subject of this book.
Whence stories?
In tackling the first question, we might begin by observing that the story cannot originate individually or collectively with the students. Their sense is of a story that is ready-made - something pre-existent and seemingly waiting for them to take up in the service of their meaning-making. We can further suggest that the story was not made up by the advertising agency that designed the billboard strat-egy. Advertising agencies, however innovative they may fancy themselves, in one respect make a virtue of being tm-original. Ads are successful when they 'hit the spot' with the target audience; a successful ad appeals to a story or stories that its target audience habitually tells itself and is therefore sympathetic to. The makers of the Kelly Browne ad were appealing to a story (about young people, parties and parents) already current in the wider social context.
It appears that there are sense-making stories that can be viewed as circulating in society, that are not easily attributable to a particular originary source. The techni-cal term for such a sense-making story is discourse. This term is variously defined (see Chapters 2 and 3). One definition regards a discourse as a coherent way of mak-ing sense of the world (or some aspect of it) as reflected in human sign systems (including verbal language). Norman Fairclough, one of the founders of CD A, draws on Michel Foucault in defining a discourse as 'a practice not just of representing the world, but of signifying the world, constituting and constructing the world in mean-ing' (1992a: 64). Questions of discourse will be explored in Chapters 2 and 3.
Critical Discourse Analysis
As we shall see, variations on the word construct (e.g. construction, constructing) are frequently found in CDA lit-erature. (Indeed, they are a reminder that CDA is itself anchored in a discourse, a way of constructing the process of meaning-making in society.) In respect of the Kelly Browne ad, we might state that it constructs young people, parties and parents in particular ways:
• Young people prefer to socialize away from their par-ents.
• Parties are occasions for behaviour which parents may well disapprove of.
• Parents are party dampeners.
Listing constructions in this way highlights the con-structedness of meaning. It also makes it easier to engage in acts of dissent - to take issue with these constructions and to resist the storied meanings any text is positioning one (another technical term) to subscribe to. When we accept a position that a text appears to be offering us or calling us to accept, we can be described as interpellated by it (a term coined by Althusser 1971). Certainly, my students were keen to emphasize the extent to which they posit-ioned themselves in opposition to the discourse that was constructing young people, parties and parents in a par-ticular way. Unsurprisingly, the dissenters were often mature students who were parents themselves and who objected to being constructed as non-partying animals.
There is another word in Fairclough's definition of discourse that bears reflecting upon because it raises another kind of issue; that is the word constitutingas in the expression * constituting the world in meaning'. Fair-clough here is drawing on a key insight of Foucault's that 'discourse is in an active relation to reality, that language signifies reality in the sense of constructing meanings for it, rather than that discourse is in a passive relation to
reality, with language merely referring to objects which are taken to be given in reality' (1992a: 41-2). The dict-ionary meaning of 'constitute' is 'to be; to go together to make'. Utilizing this definition, Fairclough's words might be rewritten: 'Discourse(s) make the world meaningful.' Or more strongly: 'Only in discourse is the world made meaningful.' There are clearly epistemological questions here, which I will simply raise. Questions such as: 'Is the world knowable outside of discourse?' and 'Can meaning-making take place outside of socially constructed sig-nifying systems?'
Finally, there is another aspect of discourse that war-rants comment. Fairclough, in his definition, also refers to discourse as a practice. While such a practice is reflected in human verbal and non-verbal signifying systems, it also embraces a range of human activities. In other words, a discourse implies ways of being and doing as well as ways of signifying. James Gee captures this aspect of discourse when he asserts that 'Discourses include much more than language':
Discourses, then, are ways of behaving, interacting, valuing, thinking, believing, speaking, and often reading and writ-ing that are accepted as instantiations of particular roles (or 'types of people') by specific groups of people, whether families of a certain sort, lawyers of a certain sort, bikers of a certain sort, business people of a certain sort, church members of a certain sort, African-Americans of a certain sort, women or men of a certain sort, and so on through a very long list. Discourses are ... 'ways of being in the world'; they are 'forms of life'. They are, thus, always and every-where social and products of social histories. (1996: viii) In the example of Kelly Browne, the notion of discourse
(Gee tends to capitalize his use of the word in this sense) extends to the roles assumed and typical activities
Critical Discourse Analysis
engaged in by people who are sympathetic to the con-structions of young people, parents and party-going that the ad invites one to subscribe to.
Acts of interpretation
The second question draws attention to the act of inter-pretation that allowed these students to come up with a coherent interpretation of such a minimal text. What is interpretation and how did it happen? Interpretation arises from an act of reading or analysis which makes meaning of a text. Extending this definition, Fairclough argues that in respect of discourse analysis, interpretation focuses on three dimensions of discursive practice: (i) its manifestation in linguistic form (in the form of 'texts'); (ii) its instantiation of a social practice (political, ideolo-gical, and so on); and (iii) a third dimension which focuses on socially constructed processes of production, distribution and consumption which determine how texts are made, circulated and used.
These dimensions will be revisited in more detail in Chapters 4-6. However, at this juncture I would point out that:
• These three dimensions are not mutually exclusive. • The first focuses on the text as a describable and
pat-terned thing made out of language but extending to other related signifying systems such as the lettering style of the Kelly Browne ad (see Chapter 2).
• The second focuses on ways in which texts reflect larger patterns of social practice - ways of identifying, ways of thinking, ways of being in the world (see Chapter 3). • The third focuses on the ways in which texts operate in
and read. It also draws attention to the relationship between texts.
The latter point can also be illustrated in reference to the Kelly Browne billboard. I argued earlier that the ad's 'story' about young people, parties and parents was circ-ulating before the ad makers decided to make use of it. The concept of intertextuality relates to ways in which texts are referenced to other texts by virtue of the stories (or discourses) embedded in them. Moreover, texts can refer forwards as well as backwards. The Kelly Browne billboard was specifically designed to link with other texts produced as part of what turned out to be an advertising campaign for an insurance company. But in ways that the billboard designers could not anticipate, the slogan was also taken up by others in 'intertextual acts'. (For example, a wag at a rugby game at Carisbrook, Dunedin, carried a placard with the words, Tarty at Tony Brown's place'. Tony Brown was the fly-half of the Otago rugby team playing that day.)
In the course of this chapter, and using a small text of nine words, we have engaged in an act of CDA. We can characterize this act as:
• analytical because we have conducted a detailed sys-tematic examination of a particular object with a view to arriving at one or more underlying principles. • discourse oriented in that this analysis has been
con-cerned with language in use (one sense of the word 'discourse') and with the way in which patterns of meaning (as in stories that make the world mean-ingful) are socially constructed (the other sense of the word 'discourse').
• critical because a central outcome of the act of analysis is to enable consideration of the social effects of the meanings a reader is being positioned or called upon
Critical Discourse Analysis
to subscribe to in the act of reading, and the
con-testation of these meanings.
The remainder of the book shows how this act can be
elaborated into a usable and potentially powerful
research method. In the next chapter, I deal with
lan-guage and its interplay with the term 'discourse'.
Language, Discourse and
Context
Language is at the heart of critical discourse analysis. But how and why? One answer relates to what is sometimes termed the 'linguistic turn' in twentieth-century thought, which has changed language from being thought of as a medium for expressing meanings that pre-exist linguistic formulation to a system that constitutes meaningfulness in its own terms. With reference to the human sciences, Parker has noted a shift — what he calls a 'turn to dis-course' - in the last 30 years 'from a notion of repre-sentation as a direct or mediated reflection of reality to a conceptual and methodological account of representat-ion as a form of significatrepresentat-ion (1999: 4-5) which actually shapes or constitutes the object denoted. Reality as pre-ceding language and shaping it has become language preceding and shaping reality. Consequently, language has now come to occupy centre stage in scholarly invest-igation.
A recognition that meaning is - even in part - socially constructed via the mediation of language and other sign systems has a number of consequences. One is a view of meaning as historically and culturally situated as opposed to being eternal, absolute and essential. A second con-sequence, especially pertinent for researchers, is a need for reflexivity and provisionality. This requires research-ers to acknowledge that particular research traditions construct the quest for and dissemination of new
knowl-Critical Discourse Analysis
edge in ways that are culturally situated and mediated by particular forms of textual practice. As Fairclough has asserted, analysis cannot be separated from interpretation and analysts need to be 'sensitive to their own inter-pretative tendencies and social reasons for them' (1992a: 35). Yet again, we are being asked to confront the ways in which our language practices constrain the how and what of our claims to know.
This chapter briefly addresses some language issues and approaches to describing language. Questions of what and how much linguistic knowledge are required to undertake CDA are addressed in Chapter 4.
Literacy: individual cognition versus social construction
Traditionally literacy is defined as the ability to read and write texts - to decode writing (as a reader) and to code language in graphic form (as a writer). In this view, text-ual interpretation is psychological - something occurring in a reader's head. If you can decode and have the necessary background information, you can understand the meaning of a text (Gee 1996: 39). Meaning, then, is something that inheres in texts and corresponds with something 'out there' in the real world. Moreover, it tends to be unitary and sharable with other competent readers.
In using the word 'traditional', I do not want to be seen as denigrating cognitive approaches to language acquisi-tion, the way human's process language, and how the mind works to make sense of the environment. Indeed, I have enormous respect for the contributions of cognitive neuroscience in this regard (for example, of Pinker (1994, 1997) and Damasio (1999)). However, the theories
of language underpinning these approaches do not articulate readily with the concerns of CD A.
An alternative to this traditional view suggests that lit-eracy be viewed not as a single thing but as a set of socially constructed practices that readers and makers of texts are apprenticed in as members of a particular social group. Not only do different types of text require different ways of reading, but the same text can also be read in different ways to generate different meanings. Textual meaning becomes multiple and therefore indeterminate. Literacy is now plural ('literacies') and characteristic of a social group's wider set of practices rather than something denoting the cognitive competence of a single individual. This sociocultural approach to literacy allows for the observation that some versions of 'literate' practice are discursively constructed as having higher status than others (Gee 1996, Pennycook 2001).
Discourse and text
Definitions of a number of terms used in this book are far from settled. A typical dictionary definition of 'discourse', for example (a formal speech or essay on a particular subject), is clearly remote from the sense(s) in which I have been using the term. Discourse analysis employs the term in two broad categories of use (Pennycook 2001, Paltridge 2000):
1 Discourse as an abstract noun denoting language in use as a social practice with particular emphasis on larger units such as paragraphs, utterances, whole texts or genres.
2 Discourse as a countable noun (one that permits pluralization) denoting a 'practice not just of representing the world, but of signifying the world,
Critical Discourse Analysis
constituting and constructing the world in meaning' (Fairclough 1992a: 64). This is Gee's (1996) 'Dis-course' with a capital 'D'.
The first of these categories relates to the concerns of this chapter; the second to Chapter 3.
Writing from a social semiotic point of view, Hodge and Kress (1988) distinguish between message, text and dis-course. The message is the smallest semiotic form, char-acterized by 'a source and a goal, a social context and purpose'. Texts and discourses are larger units. The writers distinguish between these in defining a 'text' as 'a struc-ture of messages or message traces which has a socially ascribed unity' and a 'discourse' as 'the social process in which texts are embedded'. Texts, they emphasize, have their place in a social system of signs that is dynamic. 'So texts are both the material realization of systems of signs, and also the site where change continually takes place'
(pp. 5-6).
Writing from a similar perspective, Halliday and Hasan (1985) define 'text' as 'language that is functional ... that is, doing some job in some context, as opposed to isolated words or sentences'. Referring to a text as a 'semantic unit', they distinguish two perspectives whence it can be considered. On the one hand, a text is a product, pro-duced in a particular time and place, a material artifact that can be described and analysed. On the other hand, it is a process, 'in the sense of a continuous process of semantic choice, a movement through the network of meaning potential, with each set of choices constituting the environment for a further set'. In its 'process' aspect, a text is necessarily linked to a society's linguistic system and can be thought of as an 'interactive event', a form of exchange that is dialogic in its nature (pp. 10-11). Such argumentation is designed to explain the way texts con-nect with their social context. Ideas such as interactivity
and dialogue take us to a key theorist whose ideas about language and context have contributed greatly to the theoretical underpinnings of CDA - Mikhail Bakhtin.
Bakhtin: the utterance and speech genres
'The problem with speech genres', written in 1952-3, exemplifies Bakhtin's interest in theorizing about lang-uage in use (1986). His stated focus is the diversity of 'areas of human activity', from the reading of literary texts to informal conversation, and the role language plays in these. Each of these areas is characterized by its own set of conditions and the purposes at work for its participants. Central to his view is the claim that 'Language is realized in the form of individual concrete utterances (oral and written)' that participants make (p. 60).
For Bakhtin, the conditions that prevail in a particular area of human activity are reflected in the three con-stitutive features all utterances share (see below). These features are linked to the whole of the utterance and are determined by the specific nature of what he terms the 'particular sphere of communication'. For Bakhtin, dif-ferent spheres of communication generate their own 'relatively stable types of utterances'. These types he terms 'speech genres' (p. 60), writing: 'A particular function (scientific, technical, commentarial, business, everyday) and the particular conditions of speech communication specific for each sphere give rise to particular genres, that is, certain relatively stable thematic, compositional, and stylistic types of utterances' (p. 64).
In exploring the relationship between style and genre, it is clear that the individual has not been erased in Bakhtin's thinking by a concern for the social. For him, style is individual, but genres vary in the extent to which they are conducive to reflecting individual style. However,
Critical Discourse Analysis
the social is still a determining influence. Hence the
oft-quoted sentence where Bakhtin connects language and
society: 'Utterances and their types, that is, speech genres,
are the drive belts from the history of society to the
his-tory of language' (p. 65).
The second major part of the essay focuses on the
nature of the utterance itself. In a memorable passage, he
introduces the idea of the 'organized chain of
utter-ances', which underpins the concept of intertextuality
(introduced in Chapter 1):
Moreover, any speaker is himself a respondent to a greater or lesser degree. He is not, after all, the first speaker, the one who disturbs the eternal silence of the universe. And he presupposes not only the existence of the language system he is using, but also the existence of preceding utterances -his own and others' - with which -his given utterance enters into one kind of relation or another (builds on them, polemicizes with them, or simply presumes that they are already known to the listener). Any utterance is a link in a very complexly organized chain of other utterances, (p. 69)
In a passage that anticipates Halliday and Hasan's use of
the words 'interactivity' and 'dialogic', Bakhtin writes:
'The utterance is filled with dialogic overtones, and they
must be taken into account in order to understand fully
the style of the utterance. After all, our thought itself
-philosophical, scientific, and artistic - is born and shaped
in the process of interaction and struggle with others'
thought, and this cannot but be reflected in the forms
that verbally express our thought as well' (p. 92).
Importantly, Bakhtin's dialogism is both retrospective
and anticipative. The latter he terms addressivity, and
refers to ways in which utterances are constructed to take
account of possible future responses.
as a unit of linguistic communication are determined by a
change of speaking subjects, that is, a change of speakers.
This is the first of three constitutive features of the
utterance. The second is what he calls its finalization,
which has three aspects:
1 Semantic exhaustiveness of the theme.
2 The speaker's plan or speech will.
3 Typical compositional and generic forms of
finali-zation.
The first of these we might roughly equate with content,
the second with purpose and the third with textual form or
text type. It is the last of these, which Bakhtin identifies as
most important to his purpose, and which relates to the
key notion of genre as employed in CD A. 'We speak only
in definite speech genres, that is, all our utterances have
definite and relatively typical forms of construction of the
whole. Our repertoire of oral (and written) speech genres
is rich' (p. 78).
Bakhtin's third constitutive feature is 'the relation of
the utterance to the speaker himself (the author of the
utterance) and to the other participants in speech
com-munication' (p. 84). Here, Bakhtin distinguishes two
aspects, which determine choice of linguistic means and
speech genre:
1 'the referentially semantic assignments (plan) of the
speech subject (or author)'.
2 'the expressive aspect, that is, the speaker's subjective
emotional evaluation of the referentially semantic
content of his utterance' (p. 84). As a case in point,
Bakhtin notes how a dictionary, while it may
indi-cate the stock of words that a culture has at its
dis-posal, cannot account for the way words are used in
Critical Discourse Analysis
utterances, since this is always individual and con-textual.
In conclusion, Bakhtin's concept of genre points to the typical forms of construction of an utterance. It is social in origin (in that it recalls past utterances and anticipates future ones); yet there is a clear emphasis on individual agency and creativity. While a description of a speech genre may refer to such formal features as vocabulary, syntax and structure, it is the social context that elicits them and makes them meaningful.
Social context, genre and the 'new rhetoric'
The relationship of the lexicon (dictionary) to the social context of the utterance can be thought of as exempli-fying the way in which codified sign systems in general (verbal, visual, behavioural) are rendered meaningful only in relationship to the social structures which con-stitute them. In terms of the development of CDA, the most important theorist of the text/context relationship has been M. A. K. Halliday, who developed systemic functional grammar (see Chapter 4) out of a social-semiotic perspective on language.
In the 1980s, Halliday developed a framework for describing what he termed the context of situation, the social context of a text which allowed for meaning to be exchanged.
1 The field of discourse is the general sense of what a text is about and refers to 'what is happening, to the nature of the social action that is taking place'. This aspect is comparable to Bakhtin's sphere of com-munication.
partici-pants, their relationship, their roles and relative status.
3 The mode of discourse focuses on what the language is being ask to do - its function - the way it is orga nized, the medium (print, spoken, and so on) and also 'the rhetorical mode, what is being achieved by the text in terms of such categories as persuasive, expository, didactic, and the like' (Halliday and Hasan 1985: 12).
Halliday's context of situation denoted only the immediate environment for a textual event. He intro-duced the term context of culture for the broader institut-ional and cultural environment within which the context of situation is embedded.
Halliday's colleague, Ruqaiya Hasan, in addressing questions of textual structure, used the term contextual configuration to denote the variable interrelationship between field, tenor and mode. For her, identification of a text's contextual configuration can make sense of a text's structure. It also relates to genre, which, like Bakhtin, she regards as a socialized language practice. So what is genre? Hasan states quite simply that a genre is the Verbal expression' of a contextual configuration. Specific genres (for example, expository essays, resumes, reports, various oral genres, and so on) are characterized by what Hasan calls their generic structure potential - their 'obliga-tory' elements (those that must occur); their 'optional' elements (those that can occur), the possible placement for elements and their potential for recurrence ('itera-tion'). What we have here is a potential for structural variability in the same genre but within limits (ibid.: 55-6, 64, 108).
We will be returning to Halliday and Hasan in Chapter 4. In the meantime, Halliday's use of the expression 'rhetorical mode' connects with another approach to
Critical Discourse Analysis
thinking about the relationship of text to context: the
'new rhetoric' (Andrews 1992). At its most simple,
rhetoric (in its refurbished sense) is the art of making
language work. Function is the work that language
per-forms at a particular instance in a text. A rhetorical
approach to text can be summed up in the following
points:
• People construct texts to achieve a desired result with a
particular audience.
• Textual form follows function.
• Texts are generated by contexts.
• Texts assume a social complicity between maker and
reader.
• The expectations of participants in such acts of
com-plicity become formalized in the conventions of genre.
• These conventions relate to such language features as
layout, structure, punctuation, syntax and diction.
Like other terms in this book, 'genre' signifies different
things in different approaches to the text/context
relat-ionship. Bakhtin uses it in two senses, for both the
com-plex of factors that make up the utterance as he defines it
and the 'form of construction' of the utterance as textual
product. Hasan's focus is clearly on the text as verbal
expression. Kress, in his early work, echoes Hasan in
defining genres as 'typical forms of text which link kinds
of producer, consumer, topic, medium, manner and
occasion' (Hodge and Kress 1988: 7).
Writers in the new rhetorical tradition incline to
defi-nitions of genre focusing on similarities in the context of
situation (to use Halliday's term) rather than in the text
as artefact. Freedman and Medway, for instance, define
genres as 'typical ways of engaging rhetorically with
recurring situations' (1994: 2). By the early 1990s, Kress
was viewing genre similarly, defining it as 'the
con-ventionalised aspect of the interaction' while asserting that the text in its social and cultural context was the necessary starting point for any worthwhile consideration of the forms, uses and functions of language (Kress 1993: 24). Like Freedman and Medway, he argues that it is the stability and repeatability of a social situation that leads to stability and conventionality in textual forms. These dif-fering approaches to genre need not be seen as proble-matic to CDA. Indeed, they might be seen as the complementary process/product sides of a reasonably coherent approach to textual analysis.
The following headings are useful for describing a genre: 1 Context of culture 2 Context of situation 3 Function/purpose 4 Typical content 5 Features: . layout . diction . punctuation . syntax . structure.
Box 2.1 is an example of a description of a magazine feature article.
Box 2.1 Describing a magazine feature article
1 Context of culture: magazines have a prominent and
pervasive place in Western culture. Feature articles are the 'staple' genre to be found in most magazines.
2 Context of situation: feature articles tend to be
topi-cal, dealing with issues, people and events of interest to a magazine's readers.
3 Function/purpose: feature articles fulfil a range of
functions (informing, investigating, describing, arguing for a position). To help sell magazines, they also need to be stylish, engrossing, amusing or entertaining.
4 Typical content: feature articles typically background a
topical issue, such as the profile of a prominent per-son. Depending on the magazine, there may be an emphasis on researched information or reliance on hearsay and gossip.
5 Features:
• layout: bold headlines, subheadings and sections, photographs and captions, columns, usually two fonts and text justification.
• diction: the degree of formality is affected by the pitch (i.e. the audience aimed at). Depending on pitch and content, diction may be more or less figurative, embellished, plain or colloquial.
• punctuation: in general, punctuation follows formal conventions. The presence of direct speech will necessitate speech markers (inverted commas, usually).
• syntax: again, this will be affected by the level of formality. In general, however, syntax tends to fol-low formal, correct usage with plenty of instances of subordination and coordination and cohesive devices.
• structure: here is a typical structure:
i Begins with an initial focus which sets the scene,
ii Moves to the general topic which is being written about,
iii Topic is dealt with at length, often with a variety of points of view drawn upon,
iv Article is rounded of by revisiting the initial focus.
The categorization of features used in Box 2.1 is
affected by the use of a print text example. As early as
1988, Hodge and Kress were insisting that in con-temporary society, meaning resides strongly and perva-sively in systems other than the verbal, especially the visual (p. vii). Hypertext, the medium of the Internet, is an important, multimodal signifying system. There are, I would argue, two ways in which a print genre might be 'refeatured' under pressure from the hypertextual med-ium. The first of these is in terms of feature categories. The five categories listed in terms of a print genre above start to break down when used, for example, to describe a print text with certain graphic elements. For example, in respect of a magazine display advertisement, composition is a key category of feature. Should one make this a sixth feature? Or combine it with layout under the category of 'Design'?
What categories might serve for categorizing the fea-tures of a hypertext? I would suggest:
• architecture (for three-dimensional aspects of design, for example, patterns of intra-page and extra-page linkage)
• composition (for two-dimension aspects of design, for example, the nature of the template and the relation-ship of word-based and graphic-based elements within a page)
• verbal diction and syntax (for aspects of diction and syntax within word-based elements on a site)
• graphic diction and syntax (for aspects of diction and syntax within graphic-based elements on a site)
• aural elements: voiced word language, sound effects and music
• thematic organization and cohesion (for ways in which elements on a page and between pages are organized thematically).
Critical Discourse Analysis
because changes in information and communications technology, especially under pressure from digitalization, have been profoundly affecting textual practice which, we need to keep reminding ourselves, is not just socially but also technologically mediated. One of the challenges facing CDA is to expand its theoretical repertoire for describing texts to include the new and hybrid text-types spawned by the digital revolution.
The Critical Turn: Making
Discourse Analysis Critical
The word 'critical' is a ubiquitous epithet attached to a
variety of nouns: 'critical literacy', 'critical theory',
'crit-ical approaches', 'crit'crit-ical applied linguistics', and so on.
Like other terms used in this book - 'genre' for example
- it means different things to different people. Different
discourses construct 'critical' and 'critique' in various
ways. Kincheloe and McLaren (1994), in an overview of
critical theory and qualitative research, describe as 'risky'
attempts at identifying an underlying commonality
among 'criticalists'. Nevertheless, they suggest a critical
orientation assumes:
• that all thought is fundamentally mediated by power relations that are social and historically situated
• that facts can never be isolated from the domain of values or removed from some form of ideological inscription
• that the relationship between concept and object and between signifier and signified is never stable or fixed and is often mediated by the social relations of capitalist production and consumption
• that language is central to the formation of subjectivity (conscious and unconscious unawareness)
• that certain groups in any society are privileged over others and, although the reasons for this privileging may vary widely, the oppression that characterizes con-temporary societies is most forcefully reproduced when
Critical Discourse Analysis
subordinates accept their social status as natural, neces-sary, or inevitable
• that oppression has many faces and that focusing on only one at the expense of others (e.g. class oppression versus racism) often elides the interconnections among them • that mainstream research practices are generally,
although most often unwittingly, implicated in the reproduction of systems of class, race, and gender oppression, (pp. 139-40)
I cite the above list, not so much to endorse it as to
communicate the discursive 'flavour' of the tradition that
has constructed it.
Theorists and practitioners of CDA themselves endow
the term with different shades of meaning. This is not
surprising, since criticalist traditions in general, and CDA
specifically, draw on distinct schools of social inquiry: the
neo-Marxism of the Frankfurt school, Foucaultian
archaeology, poststructuralist deconstruction and
post-modernism (see Denzin and Lincoln, 1994: 140, van Dijk
1993: 251). Acknowledging these differences, Wodak
writes that 'Basically, "critical" is to be understood as
having distance to the data, embedding the data in the
social, taking a political stance explicitly, and a focus on
self-reflection as scholars doing research' (2001: 9).
In this chapter, I discuss the 'critical' under three
headings: critique as revelation, critical practice as
self-reflexive and critical practice as socially transformative.
Rather than offering these headings as a prescription, I
suggest that they be viewed as widespread tendencies.
Indeed, there are tensions and even contradictions
between ways in which the 'critical' can be viewed and
lived.
Critique as revelation
In normal parlance, the word 'critical' denotes the habit of evaluating an object or situation in accordance with a system of rules, principles and values. I want to start with the work of Foucault, who located the 'critical' in the systematic, analytical endeavour to reveal the nature of systems of rules, principles and values as historically situated bases for critique. He called this analysis archae-ology and its product a genealogy, his key term was discourse. Foucault's essay 'Politics and the study of discourse' first appeared in the French journal Esprit in 1968. Writ-ten two years after the publication of The Order of Things: An Archaeology of the Human Sciences, this brief work afforded Foucault the opportunity of both defining his project and defending it against the charge that his approach to cultural history militates against political radicalism. At this point in his career, Foucault summed up his project as follows:
To determine, in its diverse dimensions, what the mode of existence of discourses and particularly of scientific dis-courses (their rules of formation, with their conditions, their dependencies, their transformations) must have been in Europe, since the seventeenth century, in order that the knowledge which is ours today could come to exist, and, more particularly, that knowledge which has taken as its domain this curious object which is man. (1991: 70) His project, then, was an attempt at mapping the origin and development of ways of thinking which shape mod-ern attempts to constitute human beings as objects of scientific inquiry.
Early in his essay, Foucault stresses that he is a 'plur-alist' and sets out a set of criteria for individualizing dis-courses (e.g. medicine, psychiatry, grammar), which are
Critical Discourse Analysis
plural, relative and dynamic 'entities'. (There is a distinct
echo of Bakhtin when he notes how discourses such as
sociology and psychology undergo constant changes as
new utterances - enonces - are added to them.) These
criteria are:
1 Criteria of formation. For Foucault, what
individual-izes a
discourse is 'the existence of a set of rules of
formation' for all its objects, operations, concepts
and theoretical options. There is an individualized
discursive formation whenever it is possible to
define such a set of rules' (p. 54). One
archae-ological aim, then, is to reveal the sets of rules,
which allow for a discourse to construct its field
(objects, operations, concepts, theoretical options)
in particular ways and not others.
2 Criteria of transformation or of threshold. This criteria
is concerned with discursive change. 'I shall say
natural history or psychopathology are units of
dis-course, if I can define the set of conditions which
must have been jointly fulfilled at the precise
moment of time, for it to have been possible for its
objects, operations, concepts and theoretical
options to have been formed; if I can define what
internal modifications it was capable of; finally if I
can define at what threshold of transformation new
rules of formation came into effect' (p. 54).
3 Criteria of correlation. These criteria are concerned
with what makes a discursive formation (e.g. clinical
medicine) 'autonomous' and, for Foucault, involve
the ability to define it in relation to other discourses
(e.g. biology, chemistry) and to its non-discursive
context ('institutions, social relations, economic
and political conjuncture') (p. 54).
reveal what he terms 'the episteme of a period' — not some
totalizing grand theory of the sum of its knowledge - 'but
the divergence, the distances, the oppositions, the
dif-ferences, the relations of its various scientific discourses'
(p. 55). What such analysis opens up is a space
char-acterized by an unstable and complex interplay of
dis-cursive relationships.
Later, Foucault details his interest in what makes
cer-tain statements in a field possible - 'the law of existence of
statements, that which rendered them possible - them
and none other in their place: the conditions of their
singular emergence; their correlation with other previous
or simultaneous events, discursive or otherwise' (p. 59).
The emphasis is not on the sayer (here Foucault diverges
from Bakhtin) but on the said. Hence Foucault's avowed
disinterest in the consciousness and intentionality of the
speaking subject or individual creative genius. His
archaeological method aims at the description of an
archive, by which he means 'the set of rules which at a
given period for a given society define':
1 'The limits and forms of the sayable'. This is
con-cerned with what can be said within a particular
'domain of discourse' and the form this speaking
might take. (This touches on the question of
'genre', though Foucault does not use the word
here.)
2 'The limits and forms of conservation'. This is
con-cerned with the ways in which utterances emerge,
persist, disappear and circulate.
3 The limits and forms of memory as it appears in
different discursive formations'. This concerns the
extent to which certain utterances are subscribed to
and viewed as valid at a particular time.
4 'The limits and forms of reactivation'. This is
con-cerned with ways in which older or alien discourses
Critical Discourse Analysis
are 'retained', Valued', 'imported' and
'recon-stituted'.
5 'The limits and forms of appropriation'. This is
con-cerned with who has access to what discourses. It
includes questions such as 'How is the relationship
institutionalised between the discourse, speakers
and its destined audience?' and 'How is struggle for
control of discourses conducted between classes,
nations, linguistic, cultural or ethnic collectivities?'
This last focus for definition is relevant to the third
critical tendency I discuss later in this chapter (pp.
59-60).
What does 'critical' mean in the context of Foucault's
archaeological method? In this essay, he views his
approach as critical because it poses four challenges to
traditional approaches to the history of thought.
1 The first establishes limits to the 'realm of
dis-course'. Foucault sees himself as issuing challenges:
a To the assumption that discursive
meaning-making has 'no assignable frontier'.
b To the assumption of a meaning-making
sub-ject 'which constitutes meanings and then
transcribes them into discourse' (replaced by
the subject as always discursively constituted).
c To the assumption of 'indefinitely receding
origin', which I might paraphrase as discursive
continuity in time, which Foucault replaces
with the assignation of 'thresholds and
condit-ions of birth and disappearance' (pp. 61-2).
2 The second eliminates certain binary oppositions,
for example, 'traditions and invention', old and
new, 'the dead and the living', 'the closed and the
open' and 'the static and the dynamic' (p. 62).
3 The third - and most important - is to end what
Foucault calls the denegation of discourse. In broad terms, this challenge relates to the subject matter of Chapter 2, where I described the linguistic turn as changing language (or discourse) from being thought of as a medium for expressing meanings that pre-exist linguistic formulation to a system that constitutes meaningfulness in its own terms. Fou-cault challenges three habits of mind:
a 'that of never treating discourse except as ... a simple site of expression of thoughts, imagin-ings, knowledges, unconscious themes'
b that of only seeing in discourse patterns rela-ted to the psychological traits of an author, or related to a particular style or genre, or to an idea or theme
c 'that of supposing that all operations are con-ducted prior to discourse and outside of it, in the ideality of thought or the silent gravity of practices' (pp. 62-3).
4 The fourth is to replace the uncertainties Foucault associates with cultural history, or the history of ideas with 'the analysis of discourse itself in its conditions of formation, in its serial modification, and in the play of its dependencies and correlations' with discourses being seen as 'describable' in relat-ion to other practices, including political ones
(p. 64).
Such a brief summary does scant justice to the complexity of Foucault's thought. Similar to Bakhtin, however, his rethinking of key concepts is integral to any account of the discursive formation of CDA itself. The sense he attributes to the concept of the 'critical', as he himself has indicated, allows for the highlighting of tensions and incompatibilities in other meanings of the concept, even within the same discursive formation.
Critical Discourse Analysis
Other writers on CDA have a different slant on the
object and process of revelation. This approach moves the
focus from Foucault's episteme to the individual subject
operating within particular discursive framings - what
Wodak calls 'the individual human being as a social
individual in response to available ' 'representational
resources'" (2001: 6). As this argument goes,
subscrip-tion to a particular discourse at the individual level is
likely to be an unconsciousness effect of the processes of
discursive formation that occur at the societal level.
Dis-courses are naturalized for individual subjects, who,
view-ing the world through their own discursive lenses, regard
their own position as 'common sense' rather than a
par-ticular construction of reality. Revelation occurs when
these 'common sense' positions are demystified or
denat-uralized, and exposed as discursive constructions.
In the CDA literature, this argument is often related to
a view of society as characterized by unequal power
relat-ions appearing as societal conventrelat-ions. 'Dominant
structures stabilize conventions and naturalize them, that
is, the effects of power and ideology in the production of
meaning are obscured and acquire stable and natural
forms: they are taken as "given" ' (ibid.: 3). In terms of
this view, CDA has a role in piercing the opacity of these
arrangements of structural dominance which, in van
Dijk's view, are more powerfully established via the subtle,
everyday, textual work of persuasion, dissimulation and
manipulation that sets out to change the minds of others
in one's own interests (1993: 254).
When Kincheloe and McLaren (1994) describe the
'critical project' as 'the attempt to move beyond
assimi-lated experience, the struggle to expose the way ideology
constrains the desire for self-direction, and the effort to
confront the way power reproduces itself in the
con-struction of human consciousness' (p. 152) they have in
effect distanced themselves from Foucault's distrust of
something as stable as an 'ideology' and the sense of power as originating in the agency of particular self-interested groups. They are closer to Althusser's (1971) notion of ideological state apparatus, his definition of ideology as representing 'the imaginary relationship of individuals to their real conditions of existence' (p. 153), and the way ideology functions in 'hailing' or interpellating 'concrete individuals as concrete subjects' (p. 162). For the individual subject who 'recognises' her or himself as the one who is hailed, the construction of reality embod-ied in the ideology has already achieved the status of common sense.
Certainly, 'ideology' is a concept that currently resists erasure. 'Even with differing concepts', Wodak (2001), writes, 'critical theory intends to create awareness in agents of how they are deceived about their own needs and interests' (p. 10). Again, a Foucaultian perspective would suspect the tendency in such an utterance to sug-gest a superior, critical vantage point outside of discourse. James Gee, another key figure in the CDA literature, also acknowledges the contested nature of the term, but comes up with his own definition: 'By ideology I mean a social theory which involves generalizations (beliefs, claims) about the way(s) in which goods are distributed in society' (1996: 21).
For my own part, I define an ideology is an elaborate story told about the ideal conduct of some aspect of human affairs. As I see it, its power lies in its truth value, which is determined by the number and nature of its subscription base as much as by some notion of 'explan-atory force'. In short, the truth of an ideology is deter-mined by the number subscribing to it. The related term, 'hegemony', can consequently be defined as the state of affairs which exists when the subscription base of an ideology is broad in terms of numbers and reinforced 'vertically' by the social status of its subscribers. Or to put
Critical Discourse Analysis
it more stridently, 'Hegemony is secured when the viru-lence of oppression, in its many guises (e.g. race, gender, class, sexual orientation) is accepted as consensus'
(Kincheloe and McLaren 1994: 141).
Critical practice as self-reflexive
Reading Foucault one cannot help but be stuck by a tone. Whilst an essay such as 'Politics and the study of dis-course' reveals discourses as complex, self-contradictory and unstable, there is a declarative confidence in the discoverability of the object of his analytical inquiry - the set of rules that for a particular episteme determines what can be said and not said. In his own terms, Foucault (despite his avowed distrust of genius) almost single-handedly invented a discourse - cultural archaeology, the science of discourse. For his discursive code-breaking, for revealing an underlying order of things with an emphasis on 'large-scale structures and their underlying principles' (Pennycook 2001: 31), Foucault has sometimes been termed a structuralist. (For his disavowal of such an attribution, see Foucault 1980: 111-15.)
Both postmodernity (as descriptive of a condition) and post-structuralism (as indicating a stance on the act of reading) act to undermine certainty by introducing another kind of critical tendency having a bearing on CDA as a research method. Using the idea of story dis-cussed in Chapter 1, we can think of postmodernity as a condition which erodes confidence in any single story (or grand narrative) as having superior or absolute status as an explanation for anything. Postmodernity offers a pic-ture of cultural viewpoints, discourses, 'takes' on the meaning of life, genres, jostling with one another in a kaleidoscopic melange susceptible to rapid hybridization and pastiche. In former times, its metaphor might have
been the bazaar. In these so-called 'new times' it is the
World Wide Web, where another take on things is just a
mouse click away. (For a discussion of the distinction
between postmodernism and postmodernity - or
hyper-reality - see Kincheloe and McLaren 1994: 142-3 and
Pennycook 2001. The latter describes postmodernism as
calling into question 'any claims to overarching truths
such as human nature, enlightenment, or emancipation'
and sceptical 'about talk of reality, truth, or universality'
(p. 134.))
Post-structuralism, like postmodernism, relates to the
linguistic turn (see Chapter 2) by suggesting that reality is
discursively constructed via human sign systems. One of
the thinkers associated with post-structuralism, Jacques
Derrida, invented the idea of differance to assert a
funda-mental instability in textual meaning owing to the play of
signs within language. The sense of a stable, underlying
order of things suggested by structuralism was replaced by
surface interdeterminacy and play. Deconstruction
became a procedure for demonstrating ways in which
textual meanings are actually indeterminate. Now acts of
reading are at the heart of CD A. Yet, as Patterson (1997)
points out, 'the idea that something resides in texts
awaiting extraction, or revelation, by the application of
the correct means of interpretation is precisely the
assumption that post-structuralism set out to
pro-blematise' (p. 427). Post-structuralist reading practices
acknowledge the historical situatedness of texts, gaps in
textual coherence, the indeterminacy of textual meaning
and ways in which texts encapsulate versions of reality
(Morgan 1992).
Given such powerful contemporary bases for
scepti-cism, it behoves criticalist researchers to be self-reflexive,
'to become aware of the ideological imperatives and
epistemological presuppositions that inform their
research as well as their own subjective, intersubjective,
Critical Discourse Analysis
and normative reference claims' (Kincheloe and
McLaren 1994: 140). Kincheloe and McLaren view
self-reflexivity as twofold. First, researchers need to
acknowl-edge the social constructedness of their research method.
This includes a preparedness to view the 'common sense'
meanings of the very terms used as discursively
con-structed (McLaughlin 1995). Second, researchers need to
acknowledge the provisional!ty of their findings.
Post-structuralism replaces the individual self as
orig-inary meaning-maker with the individual subjectivity as
the social product of discourse. Whether one views the
self as synonymous with subjectivity and multiple, or
dis-tinct from subjectivity and singular, is a matter of debate.
I represent the relationship between self and discourse
dialectically in Figure 3.1, while deliberately allowing for
Figure 3.1 Self and discourse
this diagram to be read in different ways. Whatever,
pro-visionality is a necessary outcome of a position which
asserts that researchers operate out of a discursively
framed and thereby contestable subjectivity. Van Dijk
(1993) writes bluntly: 'CDAis unabashedly normative: any
critique by definition presupposes an applied ethics'
(p. 253). In their overview of qualitative research and
critical theory, Kincheloe and McLaren insist that 'The empirical data derived from any study cannot be treated as simple irrefutable facts. They represent hidden assumptions - assumptions the critical researcher must dig out and expose' (1994: 144). The required researcher attitude is modesty or 'reflective humility' (ibid.: 151). And as they further argue, traditional notions of internal and external validity may need to be replaced by some-thing they call critical trustworthiness (see ibid.: 151-2 for a discussion of this).
Critical practice as socially transformative
The tendency for critical practice to be linked to a socially transformative agenda stems from a view of discourse (and ideology) as involving power relations. Earlier in this chapter, we saw how Foucault indicated that one of five things discursively constructed by an archive is 'The limits and forms of appropriation', which is concerned with who has access to what discourses. A relevant question Fouc-ault notes is: 'How is struggle for control of discourses conducted between classes, nations, linguistic, cultural or ethnic collectivities?' (1991: 60). I have suggested that the power of a discourse relates to its subscription base and the social status of its subscribers. On this basis, some discourses are more powerful than others and subscribers of non-powerful discourses are therefore marginalized and relatively disempowered.
Critical researchers tend to align themselves with a political agenda that is committed to challenging the relative power bases of competing discourses. The how of conducting this agenda and the language in which it is couched varies among critical discourse analysts. One school of thought (represented by Ruth Wodak, Teun van Dijk, Michael Meyer and others) relates this aspect of the
Critical Discourse Analysis
critical to the notion of dominance which is defined as 'the exercise of social power by elites, institutions or groups, that results in social inequality, including political, cul-tural, class, ethnic, racial and gender inequality'. Such writers view CD A as taking an 'explicit socio-political stance' and bringing about 'change through critical understanding' (van Dijk 1993: 249-50, 252-3).
Specifically, of course, CDA is concerned with the ways in which the power relations produced by discourse are maintained and/or challenged through texts and the practices which affect their production, reception and dissemination. As Wodak (2001) puts it, 'CDA takes an interest in the ways in which linguistic forms are used in various expressions and manipulations of power' (p. 11). One of the motives driving the 'Genre School' in Aust-ralia was a view that certain genres be considered genres of power and that schooling ensure that underprivileged students have access to them (Cope and Kalantzis 1993). (Critical literacy advocates in the same country critiqued the 'Genre School' for its supposed failure to recognize the discursively loaded nature of these genres.)
Both critical literacy and critical language awareness can be thought of as pedagogical approaches to literacy committed to bringing about change through critical understanding. The two approaches have much in com-mon with each other (and with critical discourse analysis) but with differing emphases depending on the theorists and practitioners themselves.
Critical literacy, as a number of commentators have pointed out (for example, Lankshear 1994, Morgan 1997) is a contested discourse that comes in various forms. Writing of the Australian variant, Morgan describes critical literacy as 'a view of language and text as always operating within and on, for or against, the inequitable socio-political arrangements of society. Cen-tral to its work therefore is the scrutiny of the linguistic
and visual forms of representation and the implicit or explicit struggle over meaning within the available sig-nifying systems' (1997: 23). As a reading practice, it tends to be underpinned by the post-structuralist principles discussed previously. Reading a text critically means developing an awareness of how texts mediate and sustain particular discourses and power relations (Lankshear 1994: 10).
Critical language awareness, as an approach to literacy education, shares these assumptions but focuses on the range of textual and non-textual practices in a society that ensure the dominance of a particular discourse (or ideology) (see Fairclough 1992b). Catherine Wallace, for example, who worked with migrant EFL students in London, focused her pedagogy on the nature of the reading practices associated with particular social groups, the influences on textual interpretation in particular contexts and the wider question: 'How is reading material produced in a particular society, that is how do texts such as newspapers, advertisements, leaflets and public inform-ation come to us in the form they do, who produces them, and how do they come to have the salience they do?' (1992: 63). Discussing the development of the Crit-ical Language Awareness Series of workbooks for students during the time when apartheid was the prevailing ideology in South Africa, Janks (1994) describes CLA as attempting to 'raise awareness of the way in which lan-guage can be used (and is used) to maintain and to challenge existing forms of power' (p. 51). So described, CLA is an overt consciousness-raising exercise, concerned to make language-users aware of the verbal and non-verbal choices that are and can be made in the production of texts and the ways in which these choices are used to reinforce particular discursive hegemonies.